GPT-4o Update: Enhanced Version Launching on October 2nd
On Wednesday, October 2nd, the default GPT-4o model will be upgraded to the latest version, gpt-4o-2024-08-06.
This updated model offers significant improvements, including a 50% reduction in input token costs and a 33% decrease in output token costs. Additionally, the latest version introduces support for Structured Outputs, enabling more efficient and precise handling of data.
For those eager to try out the new capabilities immediately, you can start using gpt-4o-2024-08-06 by specifying it as the model parameter in the API.
If you’re currently using GPT-4o as your model parameter, the system will automatically switch to the latest version on October 2nd. However, if you prefer to stick with the older version, you can do so by specifying gpt-4o-2024-05-13 as the model parameter instead.